- Email: [email protected]
Heterodiene synthesis—VI
7-mmkdrm
Vol. 27. pp. 257 10 263.
HETERODIENE
SYNTHESIS-VI’
CONFORMATIONAL STUDIES ON THE DIHYDROPYRAN RING IN 2,3-DIHYDROPYRANO[2,3-d]ISOXAZOLES, 2,3-DIHYDROPYRAN0[2,3-b]INDOLES AND 2,3-DIHYDROPYRANO[2.3-c]PYRAZOLES USING A GRAPHICAL METHOD M. J. COOK* and G. DESIMONI~~ *School
of Chemical tIstituto
Sciences, University
di Chimica
Organica
of East Anglia.
dell’Universit8.
(Received in he UK 3 August 1970; acceptedfor
Ahalract-A
graphical treatment ofcoupling
systems has afforded information structure
alter
prefer to exist in the conformer
England
publication
17 August 1970)
constant data from a number of substituted 2.3dihydropyran
about the conformational
the conformational
Norwich,
Pavia, Italy
equilibrium
signifiintly.
equilibria
within the series. Small changes in
Some trisubstituted
with all three groups in the pseudo-axial
dihydropyran
rings
position.
PRECEDING papers, various aspects of reactions of arylidene-isoxazolones, oxindoles and -pyrazolones with some vinyl ethers have been reported.2-5 The reaction gives rise to products which contain a dihydropyran ring and the configurational assignments of the substituents on the ring have been established from a consideration of coupling constants and chemical shifts in the PMR spectra.5 The assignments verified earlier conclusions obtained from chemical methods4 The variations in the magnitude of the coupling constants between rrunr protons on the saturated part of the ring indicated that not all the compounds have the same conformational preferences and this is the subject of the present investigation. Conformational studies on other unsaturated heterocyclic systems lead to the conclusion that many 6-membered ring compounds including benzodioxans6 2,3-dihydropyrans’ and the dihydropyran rings in many flavana derivatives prefer to exist in the rapidly inverting half chair forms. The sofa form appears to be limited to compounds where the unsaturated bond can conjugate with a CO group as in some flavanoneq9 or sometimes to a ring nitrogen, as in some dihydro 1,4-thiazines.‘” Clark-Lewis9 has suggested that the energy difference between half-chair and sofa forms in some flavans could be small. For the purpose of the present work we assume that the dihydropyran ring in the compounds studied here exists in the rapidly interconverting half-chair forms but similar arguments would apply to interconverting sofa forms. The coupling constants in the NMR spectra of the compounds considered (Fig 1) are abstracted from preceding papers2*’ and are listed in the Table. IN
t To whom
correspondence
during the tenure of a NATO
should be forwarded. fellowship. 257
Work
performed
at the University
of East Anglia,
hi. J. COOK and CL DE~~MOM
258
AC I: cis II: rruns
@h
III: cis IV: rrons
a: R=oNO,,R’=Me R’ = Et b: R=H, c: R=oNO,,R’=Et
v: cis VI: wuns
a: R=H,R‘=Et
a: b: c: d: e:
IX : trans-2.3~trans-3.4 X : cis-2.3~trans-3,4
I
Ph VII : VIII : a: b:
frons-2,3-rrans-3.4 cis-2,3-trans-3,4 R=H, R’ = nPr R=oNO,, R’=nPr
R=H, R’ = Et R=oNO*, R’=Et R=pOCH,R’=Et R = oNH,, R’ = Et R = oNHAcR’ = Et
a: b: c: d: e:
Rcl
R=H, R=oN02, R = pOCH,, R = oNHz, R = oNHAc,
R’ = nPr R’=nPr R’ = nPr R’ = nPr R’ = nPr
TABLE 1 cis Compds
l
____ -_-_ ___ I a 1 b I c III a Va Vb v c Vd v e Vlla VIIb IX a IX b ix c 1X d 1x e
J 23 J “MN 4,5 5-o 4.2 7.9 5-2 4.7 52 ---_4.2-___
J,,.
J,.
Jr,
trans Compds
Jti
Jmm,
J,,
A
2.5 2.8 275 25 27 25 2.5
5.0 5.3 2.8 5.4 2-8 5-4
2.8 ----
57
--
51)
_.
--
4.7 6.1 455 95 65 5.5 66 ___-6.7-_-6.3 5.95 2.5 61 2.6 6il 6.3 5.5
7.2 6,1 7.05 7.0 6.5 6-9 6.6 6.3 --.-
-
J 33
J,3,
Jm
53.4
J c* J It”” J w-m J rb __ __-_ -.__- ._._._._._. --^ --_.
II a 11 b II c IV a VI a VI b VI c VI d VI e VIIIa VIIIb X a X b
x
c
X
d
2.4 2.5 255 2.7 2.2 2.4 2.3 ____3.9_--2.5 2.15 2.1 21 20 2.0 2.15
3-I 4,8 3,3 3.3 5.9 3.7 58 4.4
----
9.85 58 805 60 9-65 5.8 9.5 61 6.95 5.6 9.0 60 62 6.2 _-__6.S____ 8.1 8.7 8.7 8.1 7.95 7.6 7.1
6.5 -.
-. ..
261
Hetcrodicnc synthesis-VI
The measured values of Ja4 can therefore be located on a straight line (line 1) arbitrarily drawn across a graph of Coupling Constant versus x where the x scale is not graduated, Fig 3. From such a graph a relative value of x of unknown magnitude, is defined for each compound. Using the same axes, the other rrans couplings J,, (in the cis series) and Jz3# (in the fruns compounds) are each plotted against their corresponding relative values of x, and this is found to give two fairly good straight lines which supports our earlier assumptions. The equations for the lines are: line 2;
J23 = ~(JM,
- J&
line 3;
Jz3. = x(Jlcjse - JZ&
+ Jzcsc + J2s3.1
(2) (3)
The slopes of the lines are found to be approximately equal but of opposite sign and this is in accord with the further assumption used below that J2,3s equals Jza3,, and J 2e3e equals and that each pair remains constant within the compounds J2c3’er studied. Since a large number of compounds are investigated, minor variations in J 2.3, (or J3,3*,). JlcJc (or Jlt3nt) a~ well as in J3rb and J3c*c from one compound to another are averaged out. By equating Eqns (2) and (3) and using the approximations that J2,3, = J2a3.a and J 2c3c J 2c3.r it is found that x equals 05 where the lines 2 and 3 intersect. Having = established the point on the x scale where x = O-5 we obtain values of +tJ3a4a + J3c4c) = 6.5 HG and itJ2a3r + J 2c3c)= 5.55 Hz; values which are compatible with those predicted from the Karplus equation.13 Furthermore, since J,, is always larger than J,,, x increases from left to right, and it is immediately clear which compounds preferentially exist in conformer a (x > 0.5) and which prefer conformer b (x < 0.5). In order to locate x = 0 or x = 1 a knowledge of any one of J3a4,, J283a, J3c4c or J 2t3t is required. Whilst none of these is known, clearly J2t3c is less than 2.8 Hz and J 3e4c is less than 2.5 Hz, the couplings found in VIIb. A consideration of other systems e.g. dioxans, l4 suggests that a value of 2 Hz for the J2e3e coupling constant is reasonable. Extrapolation of line 2 to 2 Hz on the coupling constant axis locates the value of x = 0 on the x axis; this and a graduation of the x scale is included on the graph. The present approach assumes that variations in the truns couplings is solely dependent upon the conformational equilibrium, and neglects effects arising from electronegative groups at C2. Variations in the cis couplings show that these effects may not be insignificant. Such effects may account for the scatter of some of the points about lines 2 and 3. The order of the error in the location of x = O-5 on the x scale, which arises from the approximations, can be estimated. Thus, if 0.5 H& the error in the location of J 283, J 2r3’a k 0.5 * and J2e3e = J2e3’e _+ = x = 0.5 is of the order of 004 units (using J2a3a = 9 Hz and J2c3c = 2 Hz). Further, if J 3s4, is smaller in the rrans series than in the cis series as discussed earlier, then the relative x values of the truns compounds are all fractionally too low. A”arrected line 3,” which would be drawn further to the right than at present, would displace the location of x = 05 by about 005 units for a difference of 1 Hz between J3.*, in the two series. Whilst the errors in the location of x = 0.5 are not large, error in evaluation of individual x values may be more significant. Added to the uncertainty in x = 0.5, is
M. J. COOK and 0. DISIMONI
262
the uncertainty in the choice of JlcJc to locate x = 0, plus the deviation of the compounds from line 2 or 3. For these reasons it is felt that it is not strictly proper to determine individual equilibrium constants. We prefer to consider semiquantitatively the variations in the conformational preferences within the series as a whole, for which purpose the graph is well suited. The compounds are members of a class in which small changes in structure alter the conformational equilibrium significantly. Although there are too many factors involved to enable a complete analysis of the interactions to be made, some points do clearly emerge. Most of the trans compounds prefer the form with the C,-aryl group pseudo-equatorial and the 2-alkoxy group pseudo-axial whilst the cis compounds in general have a slight preference for the form with both substituents pseudo-axial. This indicates that a 2-alkoxy group prefers the pseudo-axial position; a preference arising from dipole-dipole interactions and well known in carbohydrate chemistry (the anomeric effect),” and observed in a number of other systems also, e.g. phenanthrodioxans, l6 dioxans” and tetrahydrochromans.‘* The cis and mm indole derivatives IIIa and IVa show a stronger preference for the C,-aryl equatorial conformer than the majority of the other compounds. This suggests that the phenyl substituent on the adjacent ring in these other compounds destabilises the 4-aryl group in the pseudo-equatorial position ; an interesting result in that a small attractive force between syn periplanar 1,2diphenyl groups has been calculated for 1,bdiphenylethane.’ 9 The nature of the substituent on the C,-aryl group significantly affects the conformational equilibrium of the dihydropyran ring. This is illustrated by a comparison of the relative x value for the cis compounds Va, Vb, Vc, and Ve, and the trans compounds Via, VIb. VIc. and Vie which suggests that an ortho nitro group effectively increases the preference for the trans compounds to exist in conformer a and the cis to exist in conformer b. The reason for this is no&clear but it appears to be a function of the nitro group rather than of a generally substituted ortho position since the effect is reduced when the group is converted to NHCOMe. This phenomenon is even more pronounced in the 2,3,4_trisubstituted compound. Whereas the orrho amino, ortho acetamido and para methoxy trans tram compounds (IX, c, d and e) have no particular conformational preferences, the corresponding ortho nitro derivative IXb, as well as the analogous VIIb, probably exist almost exclusively with all three groups in the pseudo-axial positions. These latter results in particular emphasise the complexities of the interactions involved in these systems where reduced 1.3 interactions and field effects strongly modify the well known conformational rules of cyclohexane. Acknowledgements-The
authors
thanks the Consiglio Nazionale
thank
Professor
A. R. Katritzky
delle Richerche (Rome-Italy)
for his comments.
for a NATO
One of us (G.D.)
fellowship.
REFERENCES
*
: G. Tacconi,
and G. Desimoni, Tetrahedron * G. Desimoni and G. Tacconi. Communication A-30, 2nd Internotional Congress of Helerocyclic Chemistry. Montpcllier. July (1969). ’ G. Desimoni. G. Tacconi and F. Marinone. Gazz. Chim. Ital. 98. 1301 (1968) ’ G. Desimoni and G. Tacconi, Ibid. 98. 1329 (1968) ’ G. Desimoni. M. J. Cook and G. Tacconi. Ann. Chim. Rome 60.208 (1970) 6 M. J. Cook, A. R. Katriuky and M. J. Sewell, J. Chem Sot. (B). 1207 (1970) Part V
A. Gamba,
F. Marinone
Hcterodicnc synthesis--VI ’ s 9 ‘” ” I*
” I* ”
lb ”
” I9
263
D. T. Sepp and C. B. Anderson, Terrahedron 24,6873 (1968) B. J. Bolger. A. Hirwc, K. G. Marathe, E. M. Philbin. M. A. Vickers and C. P. Lillya, Ibid. 21621 (1966) J. W. Clark-Lewis, Ausrral. J. Chem 21.2059 (1968) A. R. Dunn, I. McMillan and R. J. Stoodlcy, Tefruhedron 24,2985 (1968) M. J. Cook. Tetrahedron Lerrers 2893 (1969) H. Booth Ibid. 411 (1965) M. Karplus J. Chum Phys. 30. I1 (1959) G. Gatti, A. L. Scgrc and C. Morandi Tetrahedron 23.4385 (1967) E. L. Eliel, N. L. Allinger. S. J. Angyal and G. A. Morrison. Conjiormarional Analysis. Intersciena. New York (1965) G. Pfundt and S. Farid. Tetrahedron U, 2237 (1966) C Romers, C. Altona H. R. Buys and E. Havinga, Topics in Stereochemistry 4. 39 (1969) J.‘Brugidou and H. Christol, Bull. Sot. Chim. Fr. 1974 (1966) N. Tyutyutkov and D. Petkov. CR. Acud. Bulg. Sci. 20.699 (1%7)
Vol. 27. pp. 257 10 263.
HETERODIENE
SYNTHESIS-VI’
CONFORMATIONAL STUDIES ON THE DIHYDROPYRAN RING IN 2,3-DIHYDROPYRANO[2,3-d]ISOXAZOLES, 2,3-DIHYDROPYRAN0[2,3-b]INDOLES AND 2,3-DIHYDROPYRANO[2.3-c]PYRAZOLES USING A GRAPHICAL METHOD M. J. COOK* and G. DESIMONI~~ *School
of Chemical tIstituto
Sciences, University
di Chimica
Organica
of East Anglia.
dell’Universit8.
(Received in he UK 3 August 1970; acceptedfor
Ahalract-A
graphical treatment ofcoupling
systems has afforded information structure
alter
prefer to exist in the conformer
England
publication
17 August 1970)
constant data from a number of substituted 2.3dihydropyran
about the conformational
the conformational
Norwich,
Pavia, Italy
equilibrium
signifiintly.
equilibria
within the series. Small changes in
Some trisubstituted
with all three groups in the pseudo-axial
dihydropyran
rings
position.
PRECEDING papers, various aspects of reactions of arylidene-isoxazolones, oxindoles and -pyrazolones with some vinyl ethers have been reported.2-5 The reaction gives rise to products which contain a dihydropyran ring and the configurational assignments of the substituents on the ring have been established from a consideration of coupling constants and chemical shifts in the PMR spectra.5 The assignments verified earlier conclusions obtained from chemical methods4 The variations in the magnitude of the coupling constants between rrunr protons on the saturated part of the ring indicated that not all the compounds have the same conformational preferences and this is the subject of the present investigation. Conformational studies on other unsaturated heterocyclic systems lead to the conclusion that many 6-membered ring compounds including benzodioxans6 2,3-dihydropyrans’ and the dihydropyran rings in many flavana derivatives prefer to exist in the rapidly inverting half chair forms. The sofa form appears to be limited to compounds where the unsaturated bond can conjugate with a CO group as in some flavanoneq9 or sometimes to a ring nitrogen, as in some dihydro 1,4-thiazines.‘” Clark-Lewis9 has suggested that the energy difference between half-chair and sofa forms in some flavans could be small. For the purpose of the present work we assume that the dihydropyran ring in the compounds studied here exists in the rapidly interconverting half-chair forms but similar arguments would apply to interconverting sofa forms. The coupling constants in the NMR spectra of the compounds considered (Fig 1) are abstracted from preceding papers2*’ and are listed in the Table. IN
t To whom
correspondence
during the tenure of a NATO
should be forwarded. fellowship. 257
Work
performed
at the University
of East Anglia,
hi. J. COOK and CL DE~~MOM
258
AC I: cis II: rruns
@h
III: cis IV: rrons
a: R=oNO,,R’=Me R’ = Et b: R=H, c: R=oNO,,R’=Et
v: cis VI: wuns
a: R=H,R‘=Et
a: b: c: d: e:
IX : trans-2.3~trans-3.4 X : cis-2.3~trans-3,4
I
Ph VII : VIII : a: b:
frons-2,3-rrans-3.4 cis-2,3-trans-3,4 R=H, R’ = nPr R=oNO,, R’=nPr
R=H, R’ = Et R=oNO*, R’=Et R=pOCH,R’=Et R = oNH,, R’ = Et R = oNHAcR’ = Et
a: b: c: d: e:
Rcl
R=H, R=oN02, R = pOCH,, R = oNHz, R = oNHAc,
R’ = nPr R’=nPr R’ = nPr R’ = nPr R’ = nPr
TABLE 1 cis Compds
l
____ -_-_ ___ I a 1 b I c III a Va Vb v c Vd v e Vlla VIIb IX a IX b ix c 1X d 1x e
J 23 J “MN 4,5 5-o 4.2 7.9 5-2 4.7 52 ---_4.2-___
J,,.
J,.
Jr,
trans Compds
Jti
Jmm,
J,,
A
2.5 2.8 275 25 27 25 2.5
5.0 5.3 2.8 5.4 2-8 5-4
2.8 ----
57
--
51)
_.
--
4.7 6.1 455 95 65 5.5 66 ___-6.7-_-6.3 5.95 2.5 61 2.6 6il 6.3 5.5
7.2 6,1 7.05 7.0 6.5 6-9 6.6 6.3 --.-
-
J 33
J,3,
Jm
53.4
J c* J It”” J w-m J rb __ __-_ -.__- ._._._._._. --^ --_.
II a 11 b II c IV a VI a VI b VI c VI d VI e VIIIa VIIIb X a X b
x
c
X
d
2.4 2.5 255 2.7 2.2 2.4 2.3 ____3.9_--2.5 2.15 2.1 21 20 2.0 2.15
3-I 4,8 3,3 3.3 5.9 3.7 58 4.4
----
9.85 58 805 60 9-65 5.8 9.5 61 6.95 5.6 9.0 60 62 6.2 _-__6.S____ 8.1 8.7 8.7 8.1 7.95 7.6 7.1
6.5 -.
-. ..
261
Hetcrodicnc synthesis-VI
The measured values of Ja4 can therefore be located on a straight line (line 1) arbitrarily drawn across a graph of Coupling Constant versus x where the x scale is not graduated, Fig 3. From such a graph a relative value of x of unknown magnitude, is defined for each compound. Using the same axes, the other rrans couplings J,, (in the cis series) and Jz3# (in the fruns compounds) are each plotted against their corresponding relative values of x, and this is found to give two fairly good straight lines which supports our earlier assumptions. The equations for the lines are: line 2;
J23 = ~(JM,
- J&
line 3;
Jz3. = x(Jlcjse - JZ&
+ Jzcsc + J2s3.1
(2) (3)
The slopes of the lines are found to be approximately equal but of opposite sign and this is in accord with the further assumption used below that J2,3s equals Jza3,, and J 2e3e equals and that each pair remains constant within the compounds J2c3’er studied. Since a large number of compounds are investigated, minor variations in J 2.3, (or J3,3*,). JlcJc (or Jlt3nt) a~ well as in J3rb and J3c*c from one compound to another are averaged out. By equating Eqns (2) and (3) and using the approximations that J2,3, = J2a3.a and J 2c3c J 2c3.r it is found that x equals 05 where the lines 2 and 3 intersect. Having = established the point on the x scale where x = O-5 we obtain values of +tJ3a4a + J3c4c) = 6.5 HG and itJ2a3r + J 2c3c)= 5.55 Hz; values which are compatible with those predicted from the Karplus equation.13 Furthermore, since J,, is always larger than J,,, x increases from left to right, and it is immediately clear which compounds preferentially exist in conformer a (x > 0.5) and which prefer conformer b (x < 0.5). In order to locate x = 0 or x = 1 a knowledge of any one of J3a4,, J283a, J3c4c or J 2t3t is required. Whilst none of these is known, clearly J2t3c is less than 2.8 Hz and J 3e4c is less than 2.5 Hz, the couplings found in VIIb. A consideration of other systems e.g. dioxans, l4 suggests that a value of 2 Hz for the J2e3e coupling constant is reasonable. Extrapolation of line 2 to 2 Hz on the coupling constant axis locates the value of x = 0 on the x axis; this and a graduation of the x scale is included on the graph. The present approach assumes that variations in the truns couplings is solely dependent upon the conformational equilibrium, and neglects effects arising from electronegative groups at C2. Variations in the cis couplings show that these effects may not be insignificant. Such effects may account for the scatter of some of the points about lines 2 and 3. The order of the error in the location of x = O-5 on the x scale, which arises from the approximations, can be estimated. Thus, if 0.5 H& the error in the location of J 283, J 2r3’a k 0.5 * and J2e3e = J2e3’e _+ = x = 0.5 is of the order of 004 units (using J2a3a = 9 Hz and J2c3c = 2 Hz). Further, if J 3s4, is smaller in the rrans series than in the cis series as discussed earlier, then the relative x values of the truns compounds are all fractionally too low. A”arrected line 3,” which would be drawn further to the right than at present, would displace the location of x = 05 by about 005 units for a difference of 1 Hz between J3.*, in the two series. Whilst the errors in the location of x = 0.5 are not large, error in evaluation of individual x values may be more significant. Added to the uncertainty in x = 0.5, is
M. J. COOK and 0. DISIMONI
262
the uncertainty in the choice of JlcJc to locate x = 0, plus the deviation of the compounds from line 2 or 3. For these reasons it is felt that it is not strictly proper to determine individual equilibrium constants. We prefer to consider semiquantitatively the variations in the conformational preferences within the series as a whole, for which purpose the graph is well suited. The compounds are members of a class in which small changes in structure alter the conformational equilibrium significantly. Although there are too many factors involved to enable a complete analysis of the interactions to be made, some points do clearly emerge. Most of the trans compounds prefer the form with the C,-aryl group pseudo-equatorial and the 2-alkoxy group pseudo-axial whilst the cis compounds in general have a slight preference for the form with both substituents pseudo-axial. This indicates that a 2-alkoxy group prefers the pseudo-axial position; a preference arising from dipole-dipole interactions and well known in carbohydrate chemistry (the anomeric effect),” and observed in a number of other systems also, e.g. phenanthrodioxans, l6 dioxans” and tetrahydrochromans.‘* The cis and mm indole derivatives IIIa and IVa show a stronger preference for the C,-aryl equatorial conformer than the majority of the other compounds. This suggests that the phenyl substituent on the adjacent ring in these other compounds destabilises the 4-aryl group in the pseudo-equatorial position ; an interesting result in that a small attractive force between syn periplanar 1,2diphenyl groups has been calculated for 1,bdiphenylethane.’ 9 The nature of the substituent on the C,-aryl group significantly affects the conformational equilibrium of the dihydropyran ring. This is illustrated by a comparison of the relative x value for the cis compounds Va, Vb, Vc, and Ve, and the trans compounds Via, VIb. VIc. and Vie which suggests that an ortho nitro group effectively increases the preference for the trans compounds to exist in conformer a and the cis to exist in conformer b. The reason for this is no&clear but it appears to be a function of the nitro group rather than of a generally substituted ortho position since the effect is reduced when the group is converted to NHCOMe. This phenomenon is even more pronounced in the 2,3,4_trisubstituted compound. Whereas the orrho amino, ortho acetamido and para methoxy trans tram compounds (IX, c, d and e) have no particular conformational preferences, the corresponding ortho nitro derivative IXb, as well as the analogous VIIb, probably exist almost exclusively with all three groups in the pseudo-axial positions. These latter results in particular emphasise the complexities of the interactions involved in these systems where reduced 1.3 interactions and field effects strongly modify the well known conformational rules of cyclohexane. Acknowledgements-The
authors
thanks the Consiglio Nazionale
thank
Professor
A. R. Katritzky
delle Richerche (Rome-Italy)
for his comments.
for a NATO
One of us (G.D.)
fellowship.
REFERENCES
*
: G. Tacconi,
and G. Desimoni, Tetrahedron * G. Desimoni and G. Tacconi. Communication A-30, 2nd Internotional Congress of Helerocyclic Chemistry. Montpcllier. July (1969). ’ G. Desimoni. G. Tacconi and F. Marinone. Gazz. Chim. Ital. 98. 1301 (1968) ’ G. Desimoni and G. Tacconi, Ibid. 98. 1329 (1968) ’ G. Desimoni. M. J. Cook and G. Tacconi. Ann. Chim. Rome 60.208 (1970) 6 M. J. Cook, A. R. Katriuky and M. J. Sewell, J. Chem Sot. (B). 1207 (1970) Part V
A. Gamba,
F. Marinone
Hcterodicnc synthesis--VI ’ s 9 ‘” ” I*
” I* ”
lb ”
” I9
263
D. T. Sepp and C. B. Anderson, Terrahedron 24,6873 (1968) B. J. Bolger. A. Hirwc, K. G. Marathe, E. M. Philbin. M. A. Vickers and C. P. Lillya, Ibid. 21621 (1966) J. W. Clark-Lewis, Ausrral. J. Chem 21.2059 (1968) A. R. Dunn, I. McMillan and R. J. Stoodlcy, Tefruhedron 24,2985 (1968) M. J. Cook. Tetrahedron Lerrers 2893 (1969) H. Booth Ibid. 411 (1965) M. Karplus J. Chum Phys. 30. I1 (1959) G. Gatti, A. L. Scgrc and C. Morandi Tetrahedron 23.4385 (1967) E. L. Eliel, N. L. Allinger. S. J. Angyal and G. A. Morrison. Conjiormarional Analysis. Intersciena. New York (1965) G. Pfundt and S. Farid. Tetrahedron U, 2237 (1966) C Romers, C. Altona H. R. Buys and E. Havinga, Topics in Stereochemistry 4. 39 (1969) J.‘Brugidou and H. Christol, Bull. Sot. Chim. Fr. 1974 (1966) N. Tyutyutkov and D. Petkov. CR. Acud. Bulg. Sci. 20.699 (1%7)